Maria J. Martin-Bautista

A communication model based on the 2-tuple fuzzy linguistic representation for a distributed intelligent agent system on Internet

Abstract Internet users are assisted by means of distributed intelligent agents in the information gathering process to find the fittest information to their needs. In this paper we present a distributed intelligent agent model where the communication of the evaluation of the retrieved information among the agents is carried out by using linguistic operators based on the 2-tuple fuzzy linguistic representation as a way to endow the retrieval process with a higher flexibility, uniformity and precision. The 2-tuple fuzzy linguistic representation model allows to make processes of computing with words without loss of information.

A fuzzy genetic algorithm approach to an adaptive information retrieval agent

We present an approach to a Genetic Information Retrieval Agent Filter (GIRAF) for documents from the Internet using a genetic algorithm (GA) with fuzzy set genes to learn the users information needs. The population of chromosomes with fixed length represents such users preferences. Each chromosome is associated with a fitness that may be considered the systems belief in the hypothesis that the chromosome, as a query, represents the users information needs. In a chromosome, every gene characterizes documents by a keyword and an associated occurrence frequency, represented by a certain type of a fuzzy subset of the set of positive integers. Based on the users evaluation of the documents retrieved by the chromosome, compared to the scores computed by the system, the fitness of the chromosomes is adjusted. A prototype of GIRAF has been developed and tested. The results of the test are discussed, and some directions for further works are pointed out.

Mining Text Data: Special Features and Patterns

Text mining is an increasingly important research field because of the necessity of obtaining knowledge from the enormous number of text documents available, especially on the Web. Text mining and data mining, both included in the field of information mining, are similar in some sense, and thus it may seem that data mining techniques may be adapted in a straightforward way to mine text. However, data mining deals with structured data, whereas text presents special characteristics and is basically unstructured. In this context, the aims of this paper are three: - To study particular features of text. - To identify the patterns we may look for in text. - To discuss the tools we may use for that purpose.In relation with the third point we overview existing proposals, as well as some new tools we are developing by adapting data mining tools previously developed by our research group.

Mining Fuzzy Association Rules: An Overview

The main aim of this paper is to present a revision of the most relevant results about the use of Fuzzy Sets in Data Mining, specifically in relation with the discovery of Association Rules. Fuzzy Sets Theory has been shown to be a very useful tool in Data Mining in order to represent the so-called Association Rules in a natural and human-understandable way.

Mining web documents to find additional query terms using fuzzy association rules

In this paper, we present an application of association rules to query refinement. Starting from an initial set of documents retrieved from the web, text transactions are constructed and association rules are extracted. A fuzzy extension of text transactions and association rules is employed, where the presence of the terms (items) in the documents (transactions) is determined with a value between 0 and 1. The obtained rules offer the user additional terms to be added to the query with the purpose of guiding the search and improving the retrieval.

User profiles and fuzzy logic for web retrieval issues

Abstract We present a study of the role of user profiles using fuzzy logic in web retrieval processes. Flexibility for user interaction and for adaptation in profile construction becomes an important issue. We focus our study on user profiles, including creation, modification, storage, clustering and interpretation. We also consider the role of fuzzy logic and other soft computing techniques to improve user profiles. Extended profiles contain additional information related to the user that can be used to personalize and customize the retrieval process as well as the web site. Web mining processes can be carried out by means of fuzzy clustering of these extended profiles and fuzzy rule construction. Fuzzy inference can be used in order to modify queries and extract knowledge from profiles with marketing purposes within a web framework. An architecture of a portal that could support web mining technology is also presented.

On the quest for easy-to-understand splitting rules

Decision trees are probably the most popular and commonly used classification model. They are built recursively following a top-down approach (from general concepts to particular examples) by repeated splits of the training dataset. The chosen splitting criterion may affect the accuracy of the classifier, but not significantly. In fact, none of the proposed splitting criteria in the literature has proved to be universally better than the rest. Although they all yield similar results, their complexity varies significantly, and they are not always suitable for multi-way decision trees. Here we propose two new splitting rules which obtain similar results to other well-known criteria when used to build multi-way decision trees, while their simplicity makes them ideal for non-expert users.

Textual information retrieval with user profiles using fuzzy clustering and inferencing

We present a fuzzy-logic based approach to construction and use of user profiles in web textual information retrieval. A classical user profile is a collection of terms extracted from the set of documents for a specific user or a group of users. We use a fuzzy representation for user profiles where each term in a profile is associated with a fuzzy membership value. The construction of user profiles is performed by a combination of fuzzy clustering and fuzzy inferencing, a new approach developed recently. We apply fuzzy clustering methods (such as fuzzy c-means and fuzzy hierarchical clustering) to cluster documents relevant to a user. From the cluster centers (prototypes), a user profile is constructed which indicates the users general preference of various terms. Fuzzy logic rules are also extracted from the cluster centers or from the user profiles. The fuzzy rules specify the semantic correlation among query terms. The user profiles and the fuzzy rules are subsequently used to expand user queries for better retrieval performance. Additional non-topical information about the user can be added to personalize the retrieval process. Moreover, fuzzy clustering can be applied to profiles of many users to extract knowledge about different user groups. The extracted knowledge is potentially useful for personalized marketing on the web.

Measuring Effectiveness in Fuzzy Information Retrieval

We investigate extensions of the classical measurement of effectiveness in information retrieval systems, precision and recall, to situations where the answer is modeled by a fuzzy set, such as in cases where each object in the answer is measured by its relevance to the query. The most used fuzzy extension of the classical precision-recall measure based on Zadeh’s relative cardinality appears to be counter-intuitive in some situations. We propose a new approach to the measurement of effectiveness, based on the evaluation of quantified sentences.

Text Knowledge Mining: An Alternative to Text Data Mining

In this paper we introduced an alternative view of text mining and we review several alternative views proposed by different authors. We propose a classification of text mining techniques into two main groups: techniques based on inductive inference, that we call text data mining (TDM, comprising most of the existing proposals in the literature), and techniques based on deductive or abductive inference, that we call text knowledge mining (TKM). To our knowledge, the TKM view of text mining is new though, as we shall show, several existing techniques could be considered in this group. We discuss about the possibilities and challenges of TKM techniques. We also discuss about the application of existing theories in possible future research in this field.